Abstract
Light field technologies have seen a rise in recent years and microscopy is a field where such technology has had a deep impact. The possibility to provide spatial and angular information at the same time and in a single shot brings several advantages and allows for new applications. A common goal in these applications is the calculation of a depth map to reconstruct the three-dimensional geometry of the scene. Many approaches are applicable, but most of them cannot achieve high accuracy because of the nature of such images: biological samples are usually poor in features and do not exhibit sharp colors like natural scene. Due to such conditions, standard approaches result in noisy depth maps. In this work, a robust approach is proposed where accurate depth maps can be produced exploiting the information recorded in the light field, in particular, images produced with Fourier integral Microscope. The proposed approach can be divided into three main parts. Initially, it creates two cost volumes using different focal cues, namely correspondences and defocus. Secondly, it applies filtering methods that exploit multi-scale and super-pixels cost aggregation to reduce noise and enhance the accuracy. Finally, it merges the two cost volumes and extracts a depth map through multi-label optimization.
Highlights
Light field microscopy was first introduced at Stanford in 2006 [1], and later improved in the same laboratory [2,3,4]
Local and semi-global approaches as winner-takes-all (WTA), semi-global matching (SGM) and more global matching (MGM) [40] are shown to be outperformed from global approaches, where the depth map refinement is posed as an energy minimization problem
We choose to give more weight to the correspondence matching, because it is most likely to be more accurate. We model this by creating an absolute difference map Mad = K1 |dde f,wta − dcor,wta |, where dde f,wta and dcor,wta are respectively the tentative depth map from the defocus and correspondence cost volumes and K is just a normalization factor, that will be used to weight the two contributions of each slice of the cost volume
Summary
Light field microscopy was first introduced at Stanford in 2006 [1], and later improved in the same laboratory [2,3,4]. Sensors 2019, 19, 500 of repetitive patterns that characterize microscopic samples, the task of extracting a depth map from such light fields presents many challenges To address these challenges, different methods have been proposed, whose main limitation still is the final resolution: in [17] optical flow and triangular meshes are used, and in [18] a Lytro consumer camera is used to build a light field microscope and a variational multi-scale optical flow algorithm is used to estimate the depth. The main contribution of our work is the creation of a framework where different methods are unified and adapted to obtain a more robust and accurate approach It takes the above-mentioned conditions into account and provides a method for recovering accurate depth information using a sparser light field, i.e., lower number of views with higher disparity shift. Performance Analysis to prove the quality of the proposed method is performed; in Section 5, a potential application is shown to enhance the importance of the contributions; and in Section 6, a brief summary of the proposed work is given
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
